Cafenet News

Unwiring the world

9 March 2004

There has been a lot of hot air expended on the topic of wireless hotspots with expectations of a proliferation of hotspots in public places. David Braue reports on whether the hype of public access to wireless networks in the local cafe or airport lounge is a reality and how useful it really is.

You've heard the hype about public wireless hotspots, but have you ever actually used one? Do you even know where you could find one if you wanted to?

Unless you're someone who has made it their business to find out, the answer to those questions is most likely a resounding 'no'. That's because despite all the industry hype about wireless, the rollout of Australia's nationwide Wi-Fi footprint continues at a sluggish pace that reflects the difficult commercial realities behind the model.

"In an ideal world, wireless hotspots would be so widespread that users would be able to open their devices anywhere and simply get online. This is, in theory, the goal of many hotspot providers."

A recent analysis of our Wi-Fi progress noted that Australia has just 454 public wireless LAN (PWLAN) sites, with the majority focused on airport lounges, hotel lounges, and cafes spread willy-nilly across the CBDs of Sydney and Melbourne. There's Adelaide's ambitious CitiLAN project, which provides access across much of that city's CBD through a network of 25 Wi-Fi access points, and there are around two dozen Commonwealth government-backed hotspots under trial in Brisbane and the Gold Coast.

Clearly, PWLANs are still at the experimental stage in Australia. That's a big contrast to the rest of the Asia-Pacific region, where IDC recently noted that by mid 2003 PWLAN services had attracted nearly 400,000 subscribers across the region to bring in US$13 million in revenues; this was expected to reach US$44 million and 700,000 subscribers by the end of 2003, and US$600 million and 7 million subscribers by 2007. Current annual revenue per user (ARPU) is just US$8.

Talk to competing research firm Frost & Sullivan, and you get an even more optimistic picture: the company recently projected that the number of wireless hotspot users in Australia will reach nearly 2.1 million by 2008. That's fully one-tenth the country's population, suggesting Frost & Sullivan may be a bit overenthusiastic in its eagerness to promote PWLANs. Even as wireless becomes more acceptable to consumers increasingly finding their notebooks and PDAs are shipping with Wi-Fi capabilities built in, many others are still going to be left wondering what all the hype is about.

They'll have to wait a while to find out, particularly if they don't live in the immediate vicinity of a public WLAN hotspot. Challenging financial models, still-evolving technology, an ongoing need for user education - and the screaming need for a technological solution providing seamless roaming so providers don't each have to cover the entire country - suggest that a viable model for broad public hotspot coverage will remain elusive for some time.

You can lead a horse to Wi-Fi...
In an ideal world, wireless hotspots would be so widespread that users would be able to open their devices anywhere and simply get online. This is, in theory, the goal of many hotspot providers. However, the limited range of 802.11b WLAN technology (100 metres in ideal conditions, more like 50 m in practice) means that getting good coverage requires access to lots of real estate as cheaply as possible.

Wi-Fi's early days saw hotspot providers offering financial incentives to individual businesses willing to trial the equipment for their customers, but these days many property owners are willing to let carriers install the gear on a revenue-sharing basis. Far more, however, are still reluctant to jump into the technology, leaving massive coverage holes that Wi-Fi providers have yet to fill.

Given Australia's growing cafe culture, it's not surprising that coffee bars have led the charge. Both large coffee franchises and myriad independent cafes have experimented with the technology, providing patchy coverage along the suit-covered byways of Australia's CBDs. McDonald's, which has aspirations to shed its 'fast food' image and join the cafe explosion, has also signed on, experimenting with SkyNetGlobal to roll out Wi-Fi services at its stores.

The philosophical support of massive retail chains with large numbers of physical locations is a small coup for Wi-Fi service providers, but in the grand scheme of things wireless access, it still means Australia's Wi-Fi coverage is extremely localised.

For example, stadium wireless provider VenueComm recently wired the Melbourne Cricket Ground so that photographers can instantly file their digital pictures while they're roaming around the ground - but they'd have trouble getting online across the train tracks at Melbourne Park, where IBM has set up a wireless LAN for journalists' use during its annual engagement running the technology at the Australian Open.

"In 2005, we may start to see some noticeable revenue streams, but even then I'm not sure that stream will be greater than costs."

Away from retail chains - in parks and on buildings, for example, as shown in Intel's massive Centrino advertising campaign - coverage is unpredictable at best. Companies like Xone, Azure, Telstra-backed SkyNetGlobal and Optus are all plunking down hotspots in places they think people are most likely to use them, but that's a relatively short list in most cities. Coverage will remain a significant issue for some time as Wi-Fi aspirants wrestle with the very difficult economics of trying to cover the country with patches 31,416 square metres in size.

Equipment manufacturer Netcomm hopes a bundled solution will convince more retailers to set up wireless hotspots rather than waiting for Wi-Fi providers to come knocking. Netcomm will this year offer a $2500 all-in-one WLAN kiosk that retailers can plug into a standard ADSL service. Users pay cash to get a printed receipt with a time-limited user ID and password, which they then use to log onto the access point.

Netcomm executive director Michael Boorne believes improving hotspots' ease of use will make it easier to convince retailers to take a punt on a WLAN hotspot - just as they might now try to get customers' attention by putting gumball machines, video games or cheap greeting card dispensers on the footpath in front of their shop.

"The problem with hotspots in general is that they usually ask for a credit card, and you have to read pages of terms and conditions before getting access," says Boorne. "Hotspots are a very desirable thing for people who travel a lot, but people aren't prepared to climb a mountain to get onto a hotspot for an hour. That's why we like the casual market, where buying access time becomes like buying a newspaper."

Education remains a major problem, however - customers know they can get a newspaper at any newsagent, but few know where to go for Wi-Fi access - particularly when they're away from home. Intel, among the PWLAN's staunchest proponents, recently partnered with Sensis to offer maps showing wireless hotspots through the www.whereis.com.au map search site. For its part, Azure lets users SMS a request for the nearest hotspot, then receive an SMS reply that's generated based on the phone's location.

...but you can't make it link
Just showing customers where their nearest hotspot is located isn't going to spark the Wi-Fi revolution, however. Furthermore, the WLAN industry's reliance on retailers' good graces raises another question: are fast food restaurants really suited as replacement offices? Do managers really want their field staff meeting customers at McDonald's? The chairs alone are enough to send shivers through a workplace ergonomist's spine; retail outlets simply aren't built as offices away from the office.

Where they may prove most valuable is in raising the public's awareness of hotspots. But even then, says IDC senior analyst Warren Chaisatien, the industry as a whole is going to struggle to maintain its momentum. "The industry has been operating in a loss mode for the past 18 to 24 months, and it looks set to continue that way," he says. "In 2005, we may start to see some noticeable revenue streams, but even then I'm not sure that stream will be greater than costs."

The biggest problem Chaisatien sees with the current model is its patchiness, which is accompanied by a lack of continuity between networks that each maintain their own access arrangements. One hotspot might rely on prepaid access, while its neighbour is designed to facilitate access by subscribers of a particular telecommunications carrier. Nobody would put up with having to put different types of fuel in their car to drive on different roads, but that's exactly the experience Wi-Fi providers are offering today.

Indeed, every analysis of hotspot business models focuses on the need for a seamless roaming solution to allow customers to simply turn on their systems and connect to any PWLAN hotspot, no matter which ISP is running it. This isn't possible yet, but will gradually become more real as authentication and security issues, as well as the business challenges inherent in co-operation between competitors, are worked through. ISP roaming providers like iPass and GRIC will no doubt play a role here, but their efforts at ubiquity are still in their early stages.

In the meantime, individual operators face the very real issue of building out their own PWLAN footprints in enough spaces to ensure seamless access for customers. Azure Wireless, a PWLAN wholesaler that has attracted over $50 million in funding thanks to its early entrance in the market, recently upgraded its network to around 150 hotspots with new sites in Sydney's Chifley Plaza and Chinatown, Melbourne's Crown Casino ad Chadstone Shopping Centre, and Queensland's Noosa Beach, Sanctuary Cove and Gold Coast.

"Each company's technology works over a much broader coverage area, providing wireless broadband connections that could prove irresistible to customers frustrated with hotspots' short-range coverage."

"For us, it's about offering ubiquitous coverage," says David Gold, CEO of Azure Wireless. "When we go into a venue, we tend to blanket the venue. There's no point in having wireless when people have to stay in one place."

Gold says the economic model is good enough that the company only needs two hours of usage per day, per hotspot to cover its costs. Yet one of Azure's fiercest competitors was the first to prove that the model can be profitable: in January, SkyNetGlobal posted a net profit of $792,000 for the second half of 2003. This compared favourably with a $1.62 million loss for the year-earlier period and nearly met the board's initial full-year profit prediction. Even more telling is SkyNetGlobal's reported revenue, which grew from just $963,789 in the second half of 2002 to $2.12 million during the latest reporting period.

The success of SkyNetGlobal suggests that the Wi-Fi market may well be picking up the steam it so desperately needs to become viable. Yet even as Azure and its competitors work to build out their footprints, they face threats from other emerging technologies that could knock the wind out of Wi-Fi's slowly billowing sails.

Foremost among these are iBurst and Unwired, companies that each offer wide-area wireless LAN coverage using proprietary technologies that are incompatible with Wi-Fi. Each company's technology works over a much broader coverage area, providing wireless broadband connections that could prove irresistible to customers frustrated with hotspots' short-range coverage.

Similarly, existing wireless WAN options - GPRS and 3G mobile technologies, most notably - will play an increasing role as the market comes to accept that hotspots can only be so useful. Later this year, the market will see converged devices combining GPRS and Wi-Fi capabilities into one package, with seamless roaming onto and off of faster Hi-Fi or 3G networks where they're available. This would be a reasonable compromise between coverage and speed, but will require some close partnerships between mobile network operators and Wi-Fi hotspot operators if it's going to work.

Hotspots will clearly play a role in providing localised network access in the long term. However, the difficult business model and need for complementary network options to ensure coverage will continue to produce challenges for would-be WLAN providers. Lack of profitability will no doubt see market attrition within a few years, potentially positioning PWLAN services as a value-added extension to carriers' fixed and mobile offerings. It's less glamorous than the Centrino ads, but a lot more workable in the long term.